Files
Abstract
In modern campaigns, candidates have more control over their own campaigns and they develop strategies to bring more voters and ultimately to win the election. These strategies need to be adjusted based on the changes of vote preferences as an indication of simultaneous feedback from voters during campaigns. Considering the campaigns as large-scale efforts to communicate information to voters, I focused on campaign message strategies. Candidates choose the most effective messages to reinforce and persuade voters, thereby increasing the chance of winning in the presidential election. To examine the candidates strategic behavior on selecting types of campaign messages, I categorized messages into two dimensions the positional versus valence issue dimension and the positive versus negative tone dimension. In this dissertation, I argue that candidates employ different types of campaign messages depending on the level of support from specific groups. Hypothetically, candidates deliver more positional issues to reinforce their base voters and deliver valence issues to persuade swing voters. In addition, I hypothesized that candidates deliver more positive messages to reinforce base voters and negative messages are delivered to persuade swing voters. Accordingly, the level of support from base voters and swing voters affects the decision as to which types of messages a candidate uses. To examine these dynamic campaign strategies, I selected the 1992 and 2000 presidential elections to test campaign speeches and advertisements and extended the time period to 2012 for acceptance speeches. The coded messages are regenerated as the proportional difference in the two dimensions. Daily tracking polls conducted by Gallup are used to measure the level of support from base voters and swing voters. Using the cross-sectional regression estimations, the results of this study show that campaign message strategies and the level of support are interactive. Candidates employed positional issues more to reinforce their base voters and valence issues to persuade swing voters across all types of campaign messages. Unlike this clear strategic utilization of the positional versus valence issues, the results show the mixed effects of the level of support on decisions to employ the positive or negative messages.